https://bugzilla.redhat.com/show_bug.cgi?id=1586295 --- Comment #8 from Jun Aruga <jaruga@xxxxxxxxxx> --- > I do not think wee need to do anything with the `call-to-mktemp` - see below. > > Reading the code: > * On C level [1] - expects atomic access to a file (used for caching only, reasonable hashing method is used); and > * On Ruby level [2] - Mutex is used for synchronizing the Threads Are you sure? It seems that the method "atomic_write_cache_file" which mktemp is called in is expecting as a attomic. But seeing the actual logic, I think the logic [2] is not related to the called method atomic_write_cache_file. As you know, Mutex itself is used for native thread. But it does not guarantee atomic access if lock is not used. It depends on the implementation. lib/bootsnap.rb Bootsnap::CompileCache.setup -> lib/bootsnap/compile_cache.rb Bootsnap::CompileCache.setup. -> lib/bootsnap/compile_cache/iseq.rb: Bootsnap::CompileCache::Native.fetch def self.install!(cache_dir) -> lib/bootsnap/compile_cache/yaml.rb: Bootsnap::CompileCache::Native.fetch def self.install!(cache_dir) -> ext/bootsnap/bootsnap.c rb_define_module_function(rb_mBootsnap_CompileCache_Native, "fetch", bs_rb_fetch, 3); bs_rb_fetch -> bs_fetch -> atomic_write_cache_file <= it looks atomic access from the method name. tmp_path = mktemp(template); For example it seems that when below logic is called several times at same timing, it does not guarantee the atomic access. ``` irb(main):001:0> require 'bootsnap' irb(main):006:0> Bootsnap.setup(cache_dir: '/tmp/foo', autoload_paths_cache: false) => :load_file ``` -- You are receiving this mail because: You are on the CC list for the bug. You are always notified about changes to this product and component _______________________________________________ package-review mailing list -- package-review@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe send an email to package-review-leave@xxxxxxxxxxxxxxxxxxxxxxx Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/package-review@xxxxxxxxxxxxxxxxxxxxxxx/message/3NDF7G4YLCST2LYUYIIJA3BGAG37L7MQ/
